HUMO

Examples are; is the diameter of a droplet whose volume, if multiplied by the number of droplets, equals the total volume of the sample; and Dgg (Sauter Mean Diameter, SMD) is the diameter of the drop whose ratio of volume to surface area is the same as that of the entire spray.

2.4.4.3 Representative diameters

If a distribution can be represented by a mathematical expression such as the Rosin-Rammler, it can be expressed concisely as a function of two parameters, one of which is a representative diameter and the other a measure of the spread of droplet size. Examples of representative diameter include:

Dq = Droplet diameter such that 10% of the total liquid volume is in droplets of smaller diameter

Dq 5 = Droplet diameter such that 50% of the total liquid volume is in droplets of smaller diameter. This is also known as the mass median diameter (MMD) and, if the density of the droplets is constant with size, it will also be the volume median diameter (VMD).

Dq g = Droplet diameter such that 90% of the total liquid volume is in droplets of smaller diameter.

The locations of various representative and mean diameters is shown on Figure 2.10. Various relationships between these and other representative diameters can be deduced on the basis of the mathematical expression of the distribution.

It is important to distinguish between the concept of a representative diameter and a diameter that provides an indication of atomisation quality. A reduction in the VMD does not necessarily mean that the spray is finer, instead it may be that the width of the distribution is narrower. The SMD is a better guide to the fineness of a spray and is considered by many authors (Mugele and

Evans 1951) to be the most relevant to combustion studies, although VMD is recommended as the generally used representative diameter.

24.4.4 Drop-size distribution

The width of a droplet-size distribution is an important factor when assessing the effectiveness of the atomisation process. Two sprays may have the same average droplet size but the spread of sizes may be very different. In some aspects of the study of combustion, such as particulate formation, the droplets of interest may be the very large ones present and this will not always be evident when comparing the SMD or VMD of sprays produced under different conditions. The ratio of VMD/SMD has been use by some workers (Martin and Markham 1979) to indicate the width of a distribution, but is said to be not a very sensitive indicator for some types of sprays. Another simple indicator is the relative span factor (RSP), frequently abbreviated to span, defined as;

R S P =

^ 0.5

24.4.5 Summarv of sorav characterisation

No single parameter can completely define a droplet size distribution. There is no universal correlation between the mean diameter (or representative diameter) of a spray and its droplet-size distribution. Mean diameters and representative diameters are different in nature and should not be confused. The VMD is not a mean diameter; it is a representative diameter. If a mathematical expression such as the Rosin-Rammler distribution adequately describes the observed spray, the droplet-size distribution can be defined by two parameters, a representative diameter and a measure of drop-size dispersion. The RSP can be used to indicate the spread of droplet sizes in a spray.

2.5 SUMMARY

The production of nitrogen oxides from the combustion of heavy fuel oil is affected by many variables which must be resolved in order that combustion control techniques can be made fully effective.

Some studies have shown that NgO may contribute up to over 25 % of NO^ emissions. It is important to clarify just how much NgO contributes to NO^ emissions from heavy fuel oil combustion as NgO has recently come under scrutiny as a greenhouse gas and may soon have legislative limits enforced on its emission. By continuously sampling emissions the concern over the reliability of batch sampling and subsequent NgO formation prior to analysis can be clarified.

interactions with other combustion species such as sulphur are not known. It is also unclear whether the conditions used for staged combustion for NO^ control favour NgO production. This is also true for NO^ formation which forms from rapid quenching reactions. It is therefore important to carry out work under the rich conditions used in staged combustion or other NO^- controlling techniques to investigate whether formation of NgO and NOg are favoured at the expense of a NO reduction.

Thermal NO^ formation is fairly well understood and presents few unresolved problems unlike prompt NO^ and fuel NO^ formation which still present many unanswered questions. As fuel NO^ formation is so important when considering heavy fuel oil combustion this will be the main area of discussion in this thesis. A few of the important parameters that affect fuel NO^ production are:

i) The temperature ii) The equivalence ratio iii) The fuel-nitrogen content iv) The droplet size

v) The fluid dynamics vi) Fuel properties

vii) Interaction of combustion species

The release of the fuel-nitrogen from the spray droplets is dependent on many of these parameters and is important when considering the initial environment in which the nitrogen species react, which in turn will dictate whether NO^ or Ng is formed. A thorough insight into the effect atomisation has on fuel NO^ formation is essential before many of these other parameters can be investigated. Again little information is available specifically on heavy liquid fuel combustion.

The use of emulsions for control of NO^ and particulate emissions is clear but the mechanisms behind the reductions are far from understood.

This review has shown that there are several ways in which fuel-sulphur and carbon can interact with NO^ production mechanisms. Overall the effect of sulphur is to reduce thermal NO^, but not fuel NO^, and the effect is strongest just lean of stoichiometric. As industrial combustion normally occurs with a slight excess air level the NO^-SO^ interactions must be fully understood before a fuel is desulphurised, since there is a possibility of NO^ emissions increasing in the absence of sulphur. As the majority of research has been carried out on doped gas flames and light hydrocarbons there is a need to carry out a comprehensive experimental study on a real combustion system such as heavy fuel oil. This can be carried out on the drop tube furnace (DTP) at BP which has the facility to sample at different points in the flame. The intention from

this is to produce a NO^, SO^ and carbon combustion modei which can be used in conjunction with kinetic models such as Chemkin in studying topics such as staged combustion. Only then will it be possible to establish the effect the interactions will have on legislative emissions and combustion controls.